10 Small Things That Might Be Choking Your Campaigns (14)
Affiliate Marketing is a game of tiny individual moving parts.
Just like poker, it's all about the small things that you might be doing or not doing that cause you to leak money.
Particularly if you're newer and struggling to optimise to profit, the chances are that one or more of this list will help you make the final push to profit and beyond.
Have you checked your load speed?
Loading times can kill your profit margin. But it's easy to assume that it'll be fine, or that your landers, your tracking and your offer are all smooth and swift.
It's also rarely the case. Plenty of times, the root of a failing campaign will be a massively slow tracker or lander.
You can test loading times from the US or EU with http://tools.pingdom.com - going farther afield, you can test from most countries in the world via Uptrends.org
Don't forget to test your tracker redirect speeds too - and test both your lander and your tracker under stress with a tool like JMeter or http://www.blitz.io .
Oh, and if you're on shared hosting - get off it now. NOW. You need a VPS at least - read the STM guide to web hosting to learn more.
Have you let the campaign run to statistical significance?
It's very easy to make decisions based on ads, landers or offers that just haven't had enough data to be valid.
If you're cutting a lander after 20 clicks and 2 click-throughs, or deciding that an offer's not working after 3 conversions, you're throwing money down the drain. Those 2 clickthroughs could have just been good or bad luck. You need more data to be sure.
Don't do that! We have guides to using statistical significance calculators to make sure that you're not just shooting in the dark - make sure to use them.
Likewise, don't try to test too much. If you've got a $20/day budget, don't try to test 500 ads - you'd be waiting months for data. All things being equal, testing between 6 and 10 ads and 2-3 landing pages is usually the way to go.
Have you been pausing ads, placements, and split-tests as you go?
If you're running mobile, in particular, it's vital to keep an eye on your placements and kill anything that's not performing as soon as you know for sure. You need to let them run to statistical significance, but don't let them run way beyond that.
It''s an easy, and expensive, mistake to make.
If you come to a day's traffic to optimise, but two-thirds of that traffic went to a junk placement, that's a day - and a day's budget - wasted.
Likewise, don't let your ads run too long before cutting them - particularly if you're on a single-platform web display network like POF or FB. If the ad's cost per click is just too high - if you've got a $2 payout and you're paying $1.50 per click for that ad - kill it early. Don't let your ads burn $10 or more before you kill them.
Have you thoroughly explored the bidding space?
Getting too little traffic?
Not seeing high enough profit?
You probably haven't spent enough time experimenting with bids.
If your volume's too low, have you checked to see if a massively higher bid gets you more volume? If it does, you can then experiment with steadily lower bids until you find a good balance between volume and cost.
If your costs are too high (hint - they're always too high 
), have you experimented with lower bids?
Also, have you segmented your traffic and bid differently on each segment? This is particularly crucial on mobile - you definitely shouldn't be bidding the same for App/Wifi as you are for Site/Mobile, and often customising bids even for individual large placements makes sense.
Have you asked your AM for "top volume" offers, not "top" offers?
Affiliate Managers have their own goals, and sometimes they don't align directly with yours.
If you're just asking for recommendations or top offers, they may not recommend the things that will make you the most money, but the things that will make them the most money.
Instead, ask for a list of top offers by volume, or ask specifically what volume they're doing on the offer they're recommending. That'll tell you more about whether other affiliates are having success with it.
Have you been patient?
Believe it or not, AM is not a game where you want to check and optimise your campaigns every 2 hours (except in the case of cutting placements, above).
Gather a decent amount of data over not just monetary spend, but also time, before you start making decisions and changes.
Information collected in a few hours won't be anywhere near as solid as that collected over a few days.
Set your daily budgets accordingly.
Are you jumping to conclusions?
Are you absolutely certain that you know the reason why your campaign isn't working? Then you're probably wrong.
Affiliate Marketing is all about problem-solving and research.
Before leaping to any conclusion, whether that's "this ad is awesome" or "this offer has burned out", think very, very carefully. Consider other possibilities.
What if you're wrong?
How could you tell?
Spending some time learning about problem-solving techniques, lateral thinking, and similar things will reap rewards in the long run. Look for books like "The Demon-Haunted World", "Problem Solving 101", or anything by Edward De Bono.
Even expert affiliates have to think long and hard about what areas to optimise in their campaigns.
Have you tested REALLY BIG differences?
It's very easy to think you're testing loads of ads, or loads of landers, when actually you're just testing one ad with very minor differences over and over again.
Make your ads and your landers WILDLY different from each other when you first test. Even when you're narrowing things down, be sure to include some wild-card options.
If you're just testing 3 different rules landers with differently coloured buttons, or 8 banner ads that are identical apart from one small image, you'll find it hard to improve your performance.
As a guideline - are at least some of your landers or ads utterly failing? If not, you're probably not testing widely enough. If everything's performing about the same, you need to widen your thinking.
Are you biting off more than you can chew?
If you were learning to play poker, you wouldn't do it by rocking up on the $2000 buy-in tables. Likewise, if you're learning AM, don't start by competing with the big guys.
If you're planning to do adult, for example, and your first campaign targets xHamster in the US, you're going to get crushed. That placement is targeted by massive porn companies and super-affiliates with personal deals, massive payouts and huge experience.
Likewise, if you're going for POF, don't target "all men in the US". Niche down.
Aim for smaller pockets of profit that don't have enough volume to interest the really big guys - that can still be enough to make $xxx or more a day.
Are you split-testing your offer across different networks?
It's hard to believe at first, I know, but the same offer on 2 different networks can convert up to twice as often.
If you're not testing your offer across every network you possibly can, you're probably losing quite large sums of money.
Make sure that this is the first test you do as soon as you settle on an offer, and keep doing it periodically! Just split-test the same offer against itself on as many different networks as run it.
You'll be amazed at the differences.
I hope that was helpful! If you've got any questions, comments, or suggestions for even more things that might be causing your campaigns problems, let me know below!
I needed this! Awesomesauce!
This is excellent.
A must read for guys starting out!
"Split-test across different networks" - I've seen that recommendation quite a bit but don't understand it. How can the network make a difference when everything that the user sees is exactly the same?
Originally Posted by gotepc
* The advertiser may have different scrub rates based on the relationship to an affiliate network and their overall lead quality
* Different tracking setups
* Faster / slower redirect speeds
* Affiliate Network scrubbing
* etc...
Originally Posted by gotepc
Originally Posted by gotepc
If you go into two different supermarket chain stores in the same town you would see slightly different prices - but you could think of a lot of reasons why this might happen, right?
Same thing with AM just a bit less lay-public.
I see, so is this something you would do from the start of testing, or once you know you have a successful campaign going?
Originally Posted by gotepc
Some threads just need to be printed out and kept in sight eveytime a campaign needs a fix. This is one of those!
Originally Posted by dario
It will definitely be in the next newsletter!
Great Share,
I like the most:
1.Have you asked your AM for "top volume" offers, not "top" offers?
2.Have you tested REALLY BIG differences?
And I think they are the most important
the first put you digging in the right directions, the second make you dig effectively
Home > Affiliate Marketing Forum >